Meta Prism Image Sensor Super Resolution Algorithm
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing high-resolution algorithms designed for image sensors with microlenses do not produce target high-resolution images when applied to image sensors with meta prisms, as meta prisms capture wide-angle light information differently.
Innovation Solution
An image processing device and method that processes input images from image sensors with meta prisms using a high-resolution algorithm. This involves dividing the image into sub-images, converting them to RGB demosaic images, then to YCbCr images, applying a super resolution algorithm to enhance luminance data, and merging the processed sub-images to generate an output image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a related art high-resolution algorithm designed for microlens-based image sensors is applied to meta prism-based image sensors, then the algorithm can process images from microlens sensors, but it cannot produce target high-resolution images from meta prism sensors due to different light capture characteristics
Solution Approach 1:
The patent changes the processing parameters and algorithmic approach to accommodate the different light capture characteristics of meta prism sensors. Specifically, it modifies the super-resolution algorithm to handle the wide-angle light information captured by meta prisms differently from traditional microlens-based sensors, adjusting parameters such as light path separation angles and pixel mapping relationships to achieve target resolution quality.
Solution Approach 2:
The patent segments the input image into multiple sub-images corresponding to different light paths captured by the meta prism. Each sub-image is processed separately through demosaicing and super-resolution algorithms, then the results are merged to produce the final high-resolution image. This segmentation approach allows the algorithm to handle the complex light capture pattern of meta prisms effectively.
2Measurement precision
If the input image is divided into multiple sub-images and processed through multiple conversion steps (Bayer to RGB demosaic to YCbCr), then the super resolution algorithm can effectively enhance luminance data, but the processing complexity increases
Solution Approach 1:
The patent extracts the luminance data from the color image data by converting to YCbCr color space, where the Y component represents luminance. This extraction allows the super-resolution algorithm to focus specifically on enhancing luminance data without being interfered by color information, thereby improving measurement precision while managing processing complexity through targeted processing.
Solution Approach 2:
The patent performs preliminary conversions of the input image into multiple sub-images in Bayer pattern, then converts these to RGB demosaic images and subsequently to YCbCr images before applying the super-resolution algorithm. These preliminary actions prepare the data in an optimal format for the super-resolution processing, ensuring that the luminance data is properly separated and structured before the main enhancement operation.
Data Source
AI summary
An image processing device includes an image sensor including a unit block including a plurality of pixels arranged adjacent to each other, wherein a plurality of nano-posts are arranged on the unit block, and a processor that processes an input image acquired through the image sensor.